Low-Code or Custom Software. Recalculate the Whole Cost

A low-code licence buys something different from a finished accounting product. Separate implementation cost from operation and responsibility for business rules before comparing it with custom software.

Coding tools can change an implementation estimate. Test that change against your scope rather than importing a result from another company.

A finished product and a platform carry different costs

An accounting vendor may supply statutory updates and operation, subject to the contract. An order-approval flow in Power Apps still requires your company to design its process, fields and exceptions.

The platform helps implement the workflow. It does not replace ownership of the rules. Recalculate when implementation estimates change, but do not infer savings from code-generation cost alone.

Six signs that low-code has stopped paying off

These signs justify a cost comparison. They are not a validated threshold for a rewrite.

• A consultant day on the platform costs more than a developer day.
• Adding one field means waiting for a vendor roadmap.
• Data export is markedly worse than data import.
• Somebody has a weekly routine that everyone calls "working around the system".
• The licence scales with users or records while the value delivered does not.
• The integration to your other system is a third-party connector with its own licence and its own vendor.

Put licences and connectors on the same cost sheet, alongside internal staff time.

Price statutory maintenance separately

Payroll, accounting and e-invoicing products also include ongoing attention to rules. Taking that work in-house requires an expert owner and an update budget. Cheaper implementation does not remove those duties.

A working system does not need a rewrite simply because its interface could now be generated more cheaply. Define the benefit of replacing it and the cost of reaching that benefit.

What the Shopify migration establishes

One migration can demonstrate a useful method. It does not price your project.

On 10 September 2026 Shopify announced it had rebuilt the Shop app from React Native into Swift and Kotlin in twelve weeks, with Android cold start falling from 4433 ms to 2233 ms. We went through it in detail, including what had to be in place first. What an hour of agent work costs against an hour of developer time is a separate calculation we already published.

The short version is that the agents had something to copy and a fast way to check the result. That holds on our projects too. When the new version has an existing system to diff against line by line, the work looks nothing like greenfield. Moving a process out of spreadsheets into an internal portal always brings that reference along. So does leaving a system nobody maintains.

Shopify reports its own migration results. They do not establish a general fall in development cost or the return on leaving low-code.

What you are really taking on

Include ownership after delivery in the business case.

A rewrite sells you ownership rather than code. Somebody carries the pager, and somebody has to know in three years why that exception is sitting there. A lower implementation cost can reduce the delivery budget. Long-term operating and maintenance costs still need their own estimate.

So the decision belongs in a five-year budget and not in a delivery quote. On one side sit licences, consultants and connectors, plus the hours people spend working around the thing. Against that go build, run, upgrades, plus the one person who understands it. A narrow gap is an argument for staying put. Narrow does not pay for a migration.

A clear gap is not an argument for starting with the biggest system either. Take one layer that has a reference and a check. Rewrite it, then measure whether the estimate held. It is why we open software modernization with analysis and a single bounded stage. The first stage provides local evidence for estimating the next one.

One concrete step to finish on. Pull three years of invoices for licences, for consultants, for connectors. Add them up. Use that total as one input to the comparison with a custom system.